In July 2008, the government launched the Operational Efficiency Programme (OEP), looking at where savings could be made in the UKs central government operations. Dr Martin Read was tasked with looking at back-office operations and IT, and in discussions with the government chief information officer John Suffolk came up with a set of recommendations for getting more out of the overall £16bn IT budget 4.6% of total overall central government spend. The report actually benchmarked £6.1bn of actual expenditure against three main criteria cost, user confidence and capability.

Overall, the report found that the majority of government IT operations fell within private sector norms for cost and that departments were relatively happy with what they were provided with. However, it was noted that the government was not making the most of its massive scale of purchasing, and that there was a marked duplication of functions and efforts in the provisioning and running of services as well.

In conjunction with the governments CIO Council, it was agreed that savings of around 20% can be made in the operational spend on government IT or around £3.2bn against the overall budget.

The issue is in how to do this. Over 130 datacentres were identified by the report, and work has been started in reducing these, with a view of moving towards 9-12 main government-owned datacentres. Just as in the 2004 Gershon report, the OEP report makes the recommendations of moving government services to an environment where they can be shared across departments but in the OEP case using a private cloud infrastructure where greater reuse of functions can be gained, and better economies of scale achieved. This government cloud has been termed the G-Cloud with the aim of being a simplified, standardised and available to all central set of services and functions.

All great stuff. The aim is to move to G-Cloud by 2013/14, with some services being moved into a test bed cloud during 2010. The government ICT strategy paper, launched in late January 2010, laid out the basis for how everything should move forwards.

But now a fly has appeared in the ointment: the May 6 general election. Any new government faces a major problem of reining in public sector spending. Already, the sound bites are coming to the fore, saying that government IT spend has to be controlled, as so many large projects have been such massive failures (if the headlines are to be believed).

ID cards and the NHS are the two largest projects that are currently directly threatened by all parties, but plans to cap the size of projects involving external systems integrators and the need to be seen to do something will also have an impact on the G-Cloud.

Although the Department for Work and Pensions (DWP) already has an operational cloud platform, due to physical constraints, this can only be grown to a certain level, and plans to implement G-Cloud on the scale required to meet the OEPs stated needs are now unlikely to happen within the timescales stated, due to the need for the new government to be seen to have costs completely under control.

With Suffolk having stated that the perceived problems in areas such as information security make it impossible for the government to use standard commercial cloud services, all G-Cloud eggs are currently in one, government-owned basket. If G-Cloud cannot be driven forwards at the speed needed, the savings already built in to all parties calculations will not be made and so deeper savings will have to be identified from other areas.

Six years on from Sir Peter Gershons original recommendations for shared services, the public sector is not as far along as would be expected. Dr Martin Read has identified that the same problems are still there and has made recommendations that could move towards mitigating the financial and efficiency issues that result from this.

It looks like the major opportunity to move towards a better IT platform using highly shared services on G-Cloud could be missed just so that headlines can be created that look good to the general public. If this is the case, whatever government does make active or passive decisions that stop G-Cloud from moving forwards will be doing a long-term disservice to the general public. The promise of G-Cloud has to be better communicated to the electorate, and all parties should agree that it is in the best interests of all that investment in the project is continued, so that the immense long-term benefits can be gained.

Clive Longbottom
Service Director, Business Process Analysis